How controls and suggestions divide labour

The control inventory#

Location, minimum age, language, custom exclusions: the complete fence set. Everything regulatory or commercial and non-negotiable lives here, reviewed quarterly.

Pressure-test each fence with a deliberate breach trial in a sandbox ad set: loosen one control and watch delivery spread. Trials reveal which fences truly bind. Fence trials quantify the grip.

Record fence rationale beside campaign notes: commercial non-negotiables differ from regulatory ones, and successors must know which is which. Fence records prevent accidental loosening.

Meta own audience options guide is the authority. Start at the Meta Ads hub.

Test controls deliberately: tighten one fence at a time and read delivery change. Location controls bind hardest.

The suggestion inventory#

Age, gender, interests, customs, lookalikes: the hint set that steers exploration. Suggestions shape starting points; delivery roams where signal points.

Feed suggestions with evidence: proven buyers, qualified seeds, engaged viewers. Hunches as suggestions waste the hint. Interest evidence qualifies.

Rotate suggestions as the business evolves: stale hints steer toward yesterday. Seed refreshes renew the guidance.

Score suggestion quality quarterly: proven-buyer seeds steer well while stale interest hunches steer toward yesterday. Scoring keeps hints honest. Seed scoring grades the hints.

Retire suggestions that delivery ignores: hints the system consistently roams past are noise, not guidance. Pruning sharpens exploration. Hint pruning speeds learning.

Controls in, suggestions out#

Non-negotiables in controls, hunches in suggestions: compliance, geography and brand safety fence; hypotheses hint. Every audience decision starts with this sorting question.

Document the sorting per ad set: future editors need to know what binds and what merely suggests. Setup records capture the logic.

Re-sort quarterly as hunches prove out or controls shift. Scaling reviews re-ask the question.

Run the sorting workshop with media and creative together: buyers know hunches, creatives know promises, and the divide needs both. Workshops align the account. Sorting workshops settle disputes fast.

The 2025 change map#

Consolidated options, removed exclusion types, default-on broad: 2025 reshaped the panel, so old playbooks mislead. Rebuild legacy structures on current controls rather than translating setting by setting.

Detailed-targeting exclusions became custom exclusions; detailed options consolidated. Current options verify the map.

Train every media hand on the new panel: muscle memory from 2024 builds 2026 failures. Rebuild one legacy campaign per month on current controls: gradual migration beats big-bang rebuilds that break reporting continuity. Cadence keeps risk low. Migration cadence paces the rebuild. Archive pre-migration settings with performance snapshots: rollback needs the full picture, not just the toggles. Snapshots make reversals safe. Migration archives guard the transition. Ask our team for a structures audit.

Auditing the divide#

List every control and suggestion per ad set in one sheet, then challenge each line: does this bind or merely hint, and should it? Most accounts fail this audit the first time.

Prune dead suggestions and tighten lazy controls in the same session. Overlap scoring joins the audit.

Bring finance into the audit yearly: control choices shape reach costs, and budget owners should see why fences cost what they cost. Shared audits fund better fences. Cost transparency wins budget trust.

Repeat quarterly: drift re-blurs the divide within months. Audit rhythms compound like test logs.

Controls for special cases#

Categories, teens and regulated offers lean harder on controls: where suggestions thin out, fences carry the structure. Design these accounts controls-first.

Special categories and teen limits both remove suggestion types; plan broad-with-controls deliberately. Category rules and teen limits combine.

Document compliance paths alongside performance: clean records defend accounts. Constrained playbooks thrive on fences.

Template constrained builds for reuse: category and teen accounts share patterns, so codify the controls-first scaffold once. Templates scale compliance. Compliant templates multiply the discipline.

Frequently asked questions#